Enhanced processing and analysis of multi-channel non-invasive abdominal foetal ECG signals during labor and delivery

Here the authors explore, implement and verify the potential utility of hybrid intelligent adaptive systems for processing and analysis of multi-channel non-invasive abdominal foetal electrocardiogram (fECG) signals. This approach allows clinicians to enhance non-invasive cardiotocography (CTG) with continuous ST waveform analysis (STAN) of fECG signals to improve intrapartum monitoring during labuor. The system uses a multi-channel adaptive neuro-fuzzy interference system with a new hybrid learning algorithm based on uniquely synthesised data, which comports well with real data acquired from clinical practice. The system allows the user to obtain a reference signal for objective verification. The functionality of the system has been evaluated not only by subjective criteria (an fECG morphology study by a gynaecologist), but also by objective criteria using quantitative performance metrics such as input and output signal-to-noise ratios, root mean square error, sensitivity S+, and positive predictive value among others. Experimental results indicate that hybrid neuro-fuzzy systems have the potential to improve the diagnostic and monitoring qualities (sensitivity and specificity) of fECG signals while preserving their clinically important features by leveraging the combined utility of non-invasive CTG and STAN.